Currently, a multi-user multiple-input multiple-output (MU-MIMO, Multiuser Multiple Input Multiple Output) technology is known. In MU-MIMO, multi-user transmission is implemented by using multiple parallel channels in terms of a space dimension. However, in this manner, a transmit end or a receive end needs to have all or a part of channel state information. Otherwise, the receive end cannot correctly demodulate multiple effective signals, causing a transmission failure.
With development of a Wireless Local Area Network (WLAN) technology, it is hoped that the MU-MIMO technology can be applied to a WLAN. To simultaneously schedule multiple users for channel state information feedback, when a station (Station, STA) has a limited capability, an access point (AP, Access Point) may send a null data packet announcement (NDPA, Null Data Packet Announcement), a null data packet (NDP, Null Data Packet), and a beamforming report poll (Beamforming Report Poll, BF Report Poll) frame, so that the STA performs channel estimation according to information carried in the NDPA and a high-efficiency long training sequence (HE-LTF, high efficient long training filed) carried in the NDP, and feeds back channel state information to the AP within a time period.
However, more STAs are introduced in the 802.11ax standard, and some STAs cannot immediately perform feedback. For example, the STA cannot immediately perform feedback due to a limited capability, or the AP allocates insufficient resources. In the foregoing case, the STA cannot immediately feed back channel state information, and resources may be wasted because the AP waits for a long time to receive the channel state information; consequently, channel state information transmission efficiency is reduced.